Journal

Hybrid Quantum ResNet for Time Series Classification

영문초록: Residual networks (ResNet) are known to be effective for image classi?cation. However, challenges such as computational time remain because of the signi?cant number of parameters. Quantum computing using quan- tum entanglement and quantum parallelism is an emerging computing paradigm that addresses this issue. Although quantum advantage is still studied in many research ?elds, quantum machine learning is a research area that leverages the strengths of quantum computing and machine learn- ing. In this study, we investigated the quantum speedup with respect to the number of parameters in each model for a time-series classi?cation task. This paper proposes a novel hybrid quantum residual network (HQResNet) in- spired by the classical ResNet for time-series classi?cation. HQResNet introduces a classical layer before a quantum convolutional neural network (QCNN), where the QCNN is used as a residual block. These structures enable short- cut connections and are particularly effective in achieving classi?cation tasks without a data re-uploading scheme. We used ultra-wide-band (UWB) channel impulse response data to demonstrate the performance of the proposed algo- rithm and compared the state-of-the-art benchmarks with HQResNet using evaluation metrics. The results show that HQResNet achieved high performance with a small number of trainable parameters.